Component-Enhanced Chinese Character Embeddings
This work addresses the challenge of improving NLP tasks for Chinese language processing, though it appears incremental as it adapts existing embedding concepts to Chinese characters.
The authors tackled the problem of capturing semantic information for Chinese characters by developing component-enhanced embedding models and their bigram extensions, which explore character compositions as semantic indicators, and demonstrated effectiveness with evaluations on word similarity and text classification tasks.
Distributed word representations are very useful for capturing semantic information and have been successfully applied in a variety of NLP tasks, especially on English. In this work, we innovatively develop two component-enhanced Chinese character embedding models and their bigram extensions. Distinguished from English word embeddings, our models explore the compositions of Chinese characters, which often serve as semantic indictors inherently. The evaluations on both word similarity and text classification demonstrate the effectiveness of our models.